Moeris
Ἀττικιστής

D. Sources

As noted above, Moeris often rearranges heterogenous lexicographical materials, making it difficult to securely identify his sources. In all probability, PhrynichusPhrynichus Atticista’ Eclogue is a major source, as the parallels indicated by Hansen (1998, 37–8) strongly suggest (pace Fischer 1974, 43). Moreover, those lemmas beginning with a proscribed form (see C. Content and structure) may descend from the Eclogue (see e.g. Moer. ε 4Moer. ε 4, ε 6Moer. ε 6, ε 39Moer. ε 39, λ 17Moer. λ 17, π 57Moer. π 57, τ 7Moer. τ 7, φ 16Moer. φ 16, χ 24Moer. χ 24; Hansen 1998, 38 n. 90). Affinities may, however, sometimes be due to the two lexicographers consulting a common source independently from one another, as Hansen (1998, 45–6; 56) notes. Moer. α 149–51, for example, comprises a cluster of entries sharing a very similar structure and probably taken as a block-sequence from a unique source: Moer. α 149Moer. α 149 and α 151Moer. α 151 are paralleled respectively by Phryn. Ecl. 93Phryn. Ecl. 93 and 13Phryn. Ecl. 13, while Moer. α 150Moer. α 150 has no corresponding source. This could be explained either by supposing that the source of Moer. α 150 was a now lost lemma from the Eclogue (this is possible even if we accept Fischer’s idea that the Eclogue is not abridged: on this issue, see Fischer 1974, 37; Hansen 1998, 39–40; Tribulato 2022a, 928–33), or by assuming that Moer. α 149–51 and Ecl. 13 and 93 rely on the same source. To be sure, Moeris and/or later redactors likely expanded the lexicon by adding new material from different sources. Moer. ε 7Moer. ε 7: εὐσχολῶ οὐδεὶς τῶν παλαιῶν, ἀλλὰ σχολὴν ἄγω (‘No ancient [Attic author says] εὐσχολῶ [‘to have abundant leisure’]; [one should say] σχολὴν ἄγω instead’) and Moer. ε 22Moer. ε 22: εὐκαιρεῖν οὐδεὶς τῶν παλαιῶν· Ἕλληνες δέ (‘No ancient [Attic author says] εὐκαιρεῖν [‘to have leisure’]. On the contrary, users of Greek [say so]) would appear to be two examples of this practice. As Hansen (1998, 37) notes, these entries are closely connected with Phryn. Ecl. 97Phryn. Ecl. 97: εὐκαιρεῖν οὐ λεκτέον, ἀλλ’ εὖ σχολῆς ἔχειν (‘One should not say εὐκαιρεῖν [‘to have leisure’], but rather εὖ σχολῆς ἔχειν’). It is, however, unlikely that these entries in Moeris both depend on Phrynichus: if this were indeed the case, it would be difficult to explain why Moeris should have divided Phrynichus’ text into two entries that are at a considerable distance from one another despite beginning with the same letter. Furthermore, neither of the two entries perfectly mirrors Phrynichus’ wording. Moer. ε 7 and ε 22 therefore appear to have two different sources, one of which was perhaps Phryn. Ecl. 97.

Something similar may be said about Phrynichus’ Praeparatio sophistica as a source of Moeris’ lexicon (Hansen 1998, 38–40; caution is advised, due to the abridged form of Phrynichus’ text: see entry Phrynichus Atticista, Σοφιστικὴ προπαρασκευή (Praeparatio sophistica)). One example of this is Moer. γ 23Moer. γ 23: γάργαλος Ἀττικοί· γαργαλισμός Ἕλληνες (‘Users of Attic [say] γάργαλος [‘tickling’]. Users of Greek [say] γαργαλισμός’), with its corresponding entry in Phryn. PS 56.9–10Phryn. PS 56.9–10: γάργαλος· ὁ ἐρεθισμός. καὶ γαργαλισμός. τὸ δὲ γαργαλίζεσθαι οὐκ Ἀττικόν (‘γάργαλος: [It means] irritation. [One can] also [say] γαργαλισμός. Nonetheless, γαργαλίζεσθαι [is] not Attic’). These entries are not completely equivalent: while Phrynichus’ seemingly admits of both γάργαλος and γαργαλισμός, Moeris rejects the latter as ‘Greek’. One cannot exclude that both lexicographers independently abridged a common source (see entry γάργαλος, γαργαλισμός, γαργαλίζω, γαγγαλίζω).

With regard to other lexica and erudite works, there are strong affinities between Moeris and Philemon, probably due to the fact that they often rely on Phrynichus – mutual dependence is excluded, however (see entry Philemon, Περὶ Ἀττικῆς ἀντιλογίας τῆς ἐν ταῖς λέξεσιν for further details). The parallels between Moeris and Herennius Philo/Ammonius can most likely be traced back to an undetected common source (see Wendel 1932, 2507; Hansen 1998, 51–2 states that this source conceivably based itself on works by Didymus and Tryphon). The same applies to the resemblances between the pseudo-Herodianic Philetaerus and Moeris’ lexicon (Hansen 1998, 52–4; see entry [Herodian], Φιλέταιρος (Philetaerus)).

In summary of the above discussion, Moeris often depends on Phrynichus, although some discrepancies suggest that, at times, they both independently relied on a common source. The latter must be Aelius DionysiusAelius Dionysius, and this has already been noted by Erbse (see Erbse 1950, 58, superseding Wendel 1932, 207, and entry Aelius Dionysius, Ἀττικὰ ὀνόματα; Hansen 1998, 42–7 discusses many parallels between Moeris and Aelius Dionysius). A representative case is Moer. κ 37Moer. κ 37: κεκραγμός Ἀττικοί· κραυγή Ἕλληνες (‘Users of Attic [say] κεκραγμός [‘scream’]. Users of Greek [say] κραυγή’). One can compare Phryn. Ecl. 314Phryn. Ecl. 314: κραυγασμός· παρακειμένου τοῦ κεκραγμὸς εἰπεῖν ἐρεῖ τις ἀμαθῶς κραυγασμός (‘κραυγασμός [‘scream’]: Since it is possible to say κεκραγμός, you have to be ignorant to say κραυγασμός’) and Antiatt. κ 9Antiatt. κ 9: κραυγασμός· ἀντὶ τοῦ κραυγή. Δίφιλος Ἀποβάτῃ (‘κραυγασμός: Instead of κραυγή. Diphilus in The Acrobat Horseman’). Here Moeris’ entry is a partial match for Phrynichus’ words; but the ‘Greek’ form κραυγή is only echoed in the Antiatticist (on these lemmas, see entry κεκραγμός, κραυγασμός, κράζω, κραυγάζω). As Hansen (1998, 43) points out, it is conceivable that Moeris, Phrynichus, and the Antiatticist independently reshaped a common source, which was, in all likelihood, Aelius Dionysius.

Direct dependence on Pausanias AtticistaPausanias Atticista is also plausible (see Erbse 1950, 59; Hansen 1998, 46–7; entry Pausanias, Ἀττικῶν ὀνομάτων συναγωγή), although the three examples discussed in Hansen are not decisive. Indeed, Moer. δ 27Moer. δ 27 may also be traced back to Ael.Dion. δ 35Ael.Dion. δ 35. The same holds true for Moer. δ 38Moer. δ 38, which is likely modelled on Ael.Dion. δ 5Ael.Dion. δ 5. Moer. ο 36Moer. ο 36 seems to depend on Pausanias, although this cannot be stated with certainty (see lastly Fiori 2022, 244).

In some cases, the parallels between Moeris and Aelius Dionysius or Pausanias may be a result of Moeris’ direct use of DiogenianusDiogenianus, who was one of the sources of both Aelius Dionysius and Pausanias (see Erbse 1950, 39). Some lemmas can even be traced back to Pamphilus (the author of the lexicon abridged by Diogenianus), as the numerous parallels between Moeris, Pollux, Athenaeus, Suetonius, and Galen suggest (Hansen 1998, 49). According to Hansen (1998, 49–50), it is unlikely that Moeris consulted such huge lexica as those by Pamphilus and Diogenianus, for he had at his disposal a much more suitable work for his purposes in the form of Aelius Dionysius’ lexicon. This remark is by no means conclusive, however: see e.g. Moer. θ 19Moer. θ 19 (discussed by Hansen 1998, 50). Generally speaking, Pamphilus and Diogenianus, although not specifically focused on issues of linguistic purity, may well have served as a helpful inventory of dialectal glosses.

While Hansen is sceptical of this position (Hansen 1998, 54), Homeric scholarshipHomeric scholarship may indeed have been consistently used in the lexicon, quite in keeping with the view of Homer as a native of Attica and of Homeric language as a mixture of dialects (including ‘proto-Attic’ forms): see e.g. Moer. α 4Moer. α 4, η 7Moer. η 7, κ 15Moer. κ 15, χ 19Moer. χ 19, entry ἥρῳ, entry ἱδρῶ, κυκεῶ, Ἀπόλλω, τυφῶ, entry χολάς, χόλιξ, entry ἀπολλύασιν, ἀπολλύουσιν, and other 3rd person plurals of -νύμι verbs, and E. Canon. Moer. ι 6Moer. ι 6: ἰσχίον τὸ κοῖλον τοῦ γλουτοῦ, ἐν ᾧ ἡ κοτύλη στρέφεται, Ἀττικοί· οἱ δὲ Ἕλληνες τὰ νῶτα ἰσχία (‘Users of Attic [call] ἰσχίον the hollow part of the buttock, where the κοτύλη [lit. ‘cup’] turns. Users of Greek [call] the back ἰσχία’), for example, assigns to Attic-speakers the same meaning of ἰσχίον that is also found in Homer (Hom. Il. 5.305–6: τῷ βάλεν Αἰνείαο κατ’ ἰσχίον ἔνθά τε μηρὸς | ἰσχίῳ ἐνστρέφεται, κοτύλην δέ τέ μιν καλέουσι, ‘With this he struck Aeneas on the hip, where the thigh turns in the hip joint – the cup, men call it’; transl. Murray 1924, 229). This same meaning can also be traced back to earlier scholarship on the different meanings of κοτύλη, cf. Ath. 11.479b (in a passage probably based on Pamphilus’ lexicon): κοτύλη δὲ καλεῖται καὶ ἡ τοῦ ἰσχίου κοιλότης (‘The hollow part of the hip-joint is also called κοτύλη’; transl. Olson 2009, 329, slightly adapted; note in passing that ancient scholarship discussing anatomical terms found in Homer and later poetry must partly depend on Aristophanes of ByzantiumAristophanes of Byzantium, see Ar.Byz. fr. 341 and cf. Moer. κ 55Moer. κ 55). The compiler of Moeris’ entry strongly abridged his source, apparently with a misquotation from the abovementioned Hom. Il. 5.306 (ἐν ᾧ ἡ κοτύλη στρέφεται in Moeris makes little sense and could be a mnemonic lapsus), in order to underline the Homeric usage of ἰσχίον. If so, we can conclude that rearrangements of Homeric scholarship in Moeris’ lexicon at least sometimes arise from a precise intention as opposed to casual implementation, as Hansen (1998, 54) believes.

A series of lemmas in Moeris must have originally been glosses on comedy, and it is conceivable that Moeris used a comic glossaryComic glossary as his source. Hansen (1998, 54–6), for example, parallels the comic glossary in P.Oxy. 15.1801 (= TM 63908) [2nd–3rd century CE], consisting of lemmas followed by a synonym and a locus classicus, which we also find in Moer. α 38Moer. α 38, α 88Moer. α 88, ε 51Moer. ε 51, θ 19Moer. θ 19, λ 20Moer. λ 20, and τ 5Moer. τ 5 (see also the correspondences between Moeris’ lexicon and Aristophanic scholia reported in Hansen 1998, 54–5. Moer. θ 19 and χ 37Moer. χ 37, however, are probably not explanations of Eubulus and Eupolis, as Hansen 1998, 54 n. 135 believes). According to Hansen, such glossaries were prone to be incorporated in the dualistic schema of Moeris’ lexicon. As a matter of fact, a number of entries that clearly rely on scholarship on comedy show no trace at all of any contraposition between Attic-speakers and Greek-speakers (see e.g. Moer. κ 32Moer. κ 32 and φ 33Moer. φ 33).

A number of lemmas dealing with Realien and ‘juridical’ termsLegal language are paralleled by Pollux, Harpocration, and ‘Bekker’s 5th lexicon’ (cf. also Platonic exegesis, on which see e.g. Valente 2012, 65). Such affinities may be traced back to that ‘Attic onomasticon’ (Wentzel 1895, 483; cf. Valente 2012, 65 n. 274) which, according to Alpers (1981, 117–23), was later partially alphabetised and expanded by Julian in his lexicon (described in Phot. Bibl. cod. 150.99a–b; see also Ucciardello 2012, 42–3; Musino 2018, 404, 409, 419–20; Matthaios 2020, 359–61). Interestingly, some of these entries do not involve the usual dualistic schema. Instead, they define a single word or expression with no apparent prescriptive intent (see e.g. Moer. α 166Moer. α 166, β 31Moer. β 31, γ 1Moer. γ 1, ε 3Moer. ε 3, ε 14Moer. ε 14, λ 4Moer. λ 4, and χ 20Moer. χ 20). Conceivably, Moeris may rely on an intermediate source, possibly Aelius Dionysius and/or Pausanias (see e.g. Moer. α 147Moer. α 147 and Ael.Dion. α 18Ael.Dion. α 18, and note that Aelius collected, among others, lemmas referring to Attic ‘juridical’ expressions; see entry Aelius Dionysius, Ἀττικὰ ὀνόματα). If so, new conjectural fragments of Aelius Dionysius and Pausanias may be detected on the basis of the agreements between Moeris and other scholarly works ultimately resting on the now lost ‘Attic onomasticon’, even when no traces of Aelius’ and/or Pausanias’ discussions on a particular subject are otherwise extant.

As illustrated above, therefore, Moeris’ sources are likely to have been Phrynichus, Aelius Dionysius, and Pausanias. It is possible, although far from certain, that Moeris consulted the lexica of Pamphilus and Diogenianus. He may also have had a comic glossary at his disposal. Then again, affinities with Philemon, Herennius Philo/Ammonius, and the Ps.-Herodianic Philetaerus must arise from them sharing a common source. Homeric scholarship is attested as well, although often strongly reshaped. The same holds true for lexicographical material that ultimately depends on that ‘Attic onomasticon’ which, according to Alpers, survived in alphabetised and expanded form in Julian’s lexicon.

E. Canon

Since Moeris’ lexicon was most likely rearranged, the task of better defining Moeris’ canon cannot simply rely on the surviving quotations from canonical authors; rather, it needs to be largely based on indirect evidence, a task made difficult by Hansen’s deficient apparatus of loci classici (on account of this, a new edition of the lexicon would be welcome). Be that as it may, the following authors are explicitly mentioned in the extant version of the lexicon: Plato (24x; 6x without reference to works), Aristophanes (17x; 6x without reference to the plays), Thucydides (16x), Xenophon (7x; 3x without reference to works), Demosthenes (6x; Demosthenes’ mention in Moer. o 33Moer. ο 33 is missing in Hansen’s index). Homer, Herodotus, Eupolis, and Plato Comicus each receive two mentions, while Euripides, Hermippus, Eubulus, Antiphon, Hypereides, Isaeus, Isocrates (Moer. κ 51Moer. κ 51: MSS wrongly read Hippocrates), Theopompus, and Aristotle are each cited only once. Given that Eubulus, Aristotle, and Theopompus only serve as parallels illustrating a specific locus classicus (see Wendel 1932, 2503) and that tragic authors are virtually absent (Euripides’ mention raises problems; see below), Moeris’ canon is comparable to that of other strict Atticists (see entry Phrynichus Atticista, Σοφιστικὴ προπαρασκευή (Praeparatio sophistica)). Indeed, this comparison is further supported by the fact that Moeris often discards Xenophon’s authority as unreliable (see below and Favi 2022, 320 n. 45).

Some features call for further explanation, however. Homeric material is consistently used, as noted above (D. Sources). This is hardly due to a ‘mistake’ (as Hansen 1998, 54 believes). Instead, as Wendel (1932, 2503–4) recognised long ago, the presence of Homeric material in Moeris is a ‘consequence of the prominence of the Homeric linguistic model’ (my translation). The scholarly tradition that regards HomerHomer as a native of Attica may have also played a role: see Swain (1996, 56), entry ἥρῳ, entry οἶσε, entry ἱδρῶ, κυκεῶ, Ἀπόλλω, τυφῶ, entry χολάς, χόλιξ, and entry ἀπολλύασιν, ἀπολλύουσιν, and other 3rd person plurals of -νύμι verbs. That said, we can detect a sporadic refusal of Homeric forms, which perhaps depends on the sources Moeris consulted. A case in point is Moer. ε 50Moer. ε 50: ἐρυγγάνων <Ἀττικοί>· ἐρευγόμενος <Ἕλληνες> (‘<Users of Attic> [say] ἐρυγγάνων [‘belching’]. Users of Greek [say] ἐρευγόμενος’), which can be compared with Phryn. Ecl. 42Phryn. Ecl. 42: ἐρεύγεσθαι ὁ ποιητής· ‘ὁ δ’ ἐρεύγετο οἰνοβαρείων’, ἀλλ’ ὁ πολιτικὸς ἐρυγγάνειν λεγέτω (‘The poet [i.e. Homer] [says] ἐρεύγεσθαι: ‘and he vomited [ἐρεύγετο] in his drunken sleep’ [Hom. Od. 9.374; transl. Murray 1919, 343]. The urbane [rhetor] should say ἐρυγγάνειν, however’; on the issues raised by these entries, see entry ἐρυγγάνω, ἤρυγον, ἐρεύγομαι, ἠρευξάμην; on the meaning of πολιτικός in Phrynichus’ works, see entry ἄπαρνος, ἔξαρνος and entry ἄψοφον ἔχειν στόμα).

As for Herodotus, in Moer. λ 18Moer. λ 18 the Herodotean usage is associated with that of users of Attic, while in Moer. θ 12Moer. θ 12, by contrast, it is compared with that of Greek-speakers. Indeed, Herodotus was considered to be an important stylistic and linguistic model, so it is not surprising to find him cited as a touchstone on such matters. It is possible that Moeris’ apparent inconsistency reflects the view that Herodotus’s language was mixed (see Tribulato 2022b, 245–6, with further references), but evidence is too scanty to draw any plausible conclusions.

Tragic poets are not cited, with the exception of Moer. ψ 13Moer. ψ 13: ψήκτρα τὸ ψηκτρίον, ὥσπερ Εὐριπίδης ἐν Φαίδρᾳ· ‘πρώην μὲν ἵππους ἐκτενίζομεν ψήκτραις’ (‘ψήκτρα [means] ψηκτρίον, as Euripides [says] in Phaedra [actually Hipp. 1174]: ‘Yesterday we curried the horses with the curry-combs’’). This lemma is only reported by cod. E (which contains a strongly epitomisedEpitome version of Moeris’ lexicon; see Hansen 1998, 32–3) and we cannot discount the possibility that it is a later addition; we may note in passing that both the title and the text of the locus classicus are misquoted, probably due to a slip of memory. It is also notable that Poll. 10.55Poll. 10.55 quotes Ar. fr. 66 as the locus classicus for ψήκτρα, which would have been much more in line with Moeris’ inclination towards Old Comedy (cf. also Phryn. PS 127.17–8Phryn. PS 127.17–8; Pollux also quotes Soph. fr. 475, however).

Hierarchies of the canonical authors can be detected here and there in the lexicon. Moeris apparently shows a preference for Thucydides and Plato when he chooses between these and Demosthenes: see entry ἀδυναμία, ἀδυνασία, ἀδυνατία; Moer. β 10Moer. β 10: βιβλία διὰ τοῦ ι, ὡς Πλάτων, Ἀττικοί· βυβλία, ὡς Δημοσθένης, κοινόν (‘Users of Attic [say] βιβλία, with ι, as Plato. βυβλία, as Demosthenes [says], [is] common’); Moer. μ 3Moer. μ 3: μετέπεμψε Θουκυδίδης· μετεπέμψατο Δημοσθένης (‘Thucydides [says] μετέπεμψε. Demosthenes [says] μετεπέμψατο’; cf. Maidhof 1912, 306–7). See also Moer. ο 33Moer. ο 33.

Moeris also prefers Thucydides over Antiphon, at least judging from Moer. λ 27Moer. λ 27: λιθουργούς Θουκυδίδης· λιθοκόπους Ἀντιφῶν (‘Thucydides [says] λιθουργούς [‘stone-masons’, acc. pl.]. Antiphon [says] λιθοκόπους’); cf. Thom.Mag. 221.6–7: τὸ δὲ λιθοκόπος ἀδόκιμον, εἰ καὶ Ἀντιφῶν λέγει (‘λιθοκόπος is unapproved, even if Antiphon uses it’). In this last case, however, we are not compelled to consider Antiphon to be less eligible as a model than Thucydides, because Antiphon was prominent in the canon of the ten Attic orators. Rather, the extreme rarity of λιθοκόπος (a quasi-hapax) may have led Moeris to reject this term in favour of λιθουργός, which has good Attic pedigree (it also appears in Ar. Av. 1134).

In summary, it is conceivable that Moeris generally gave prominence to Thucydides, Plato, and Aristophanes: this would be in line with those entries that prescribe ‘old Attic’Attic, old forms (see F. Evaluative terminology). Nevertheless, not all forms that occur in the canonical authors receive uncritical approval: those that are sporadically attested are apparently treated with caution, as is the case in Phrynichus’ Praeparatio sophistica (see Tribulato 2024, and entry Phrynichus Atticista, Σοφιστικὴ προπαρασκευή (Praeparatio sophistica)). A case in point is Moer. ε 39Moer. ε 39 (= Phryn. PS fr. *313Phryn. PS fr. *313): εὔμορφον παρ’ οὐδενὶ εὗρον, ἀλλὰ εὐπρεπῆ. εὐμορφίαν παρὰ Πλάτωνι ἐν Νόμοις καὶ παρὰ Ξενοφῶντι, σπάνιον δέ (‘I did not find εὔμορφον [‘fair of form’] in any [author], but [I found] εὐπρεπῆ instead. In Plato’s Laws [716a.6; 744c.1] and in Xenophon [Smp. 8.6] [there is] εὐμορφίαν [‘beauty of form’], but [it is] rare’). See also Moer. σ 25Moer. σ 25: συγγενήσεσιν ἀντὶ τοῦ συνουσίαις (so cod. C; Hansen mistakenly wrote -ίας) Πλάτων Νόμοις. παρ’ ἄλλῳ οὐχ εὗρον (‘Plato in Laws [948e.3–4] [says] συγγενήσεσιν [‘meetings’, dat. pl.] instead of συνουσίαις. I did not find [it] in any other [author]’).

Other cases are more problematic, however. Sometimes a form is simply ascribed to an author without any further evaluation. It is possible that not all entries of this kind had a prescriptive intent, and a case-by-case study would be necessary to show that. I give the following as examples:

• Moer. χ 35Moer. χ 35: χάριν σοι ἔχω ἀντὶ τοῦ οἶδα Πλάτων Πρωταγόρᾳ (‘Plato in Protagoras [328d.8] [says] χάριν σοι ἔχω [‘I owe gratitude to you’] instead of [χάριν] οἶδα’). Plato uses both expressions, even if χάριν ἔχειν occurs more often: generally speaking, χάριν ἔχειν is by far the most common form in Greek. It is not likely that Moeris is prescribing an expression here that could be perceived as more trivial than the other. In this regard, note that [Hdn.] Philet. 264[Hdn.] Philet. 264 prescribes both forms (Phryn. Ecl. 10Phryn. Ecl. 10 is not a perfect parallel to Moeris’ entry). It is conceivable that the two entries are based on the same source (see D. Sources). The latter likely had no prescriptive intent, and was instead only concerned with Platonic vocabulary.

• Moer. ω 9Moer. ω 9: ὡς αὕτως ἀντὶ τοῦ παραπλησίως καὶ κατὰ ταὐτά Πλάτων Ἱππίᾳ βʹ (‘Plato in Hippias Minor [371e.3; 374e.6] [says] ὡς αὕτως [‘in like manner’, ‘just so’] instead of παραπλησίως and κατὰ ταὐτά’). Plato uses ὡς αὕτως, παραπλησίως, and κατὰ ταὐτά to mean roughly the same thing; cf. e.g. Pl. Plt. 269d.5, Pl. Phd. 78d.8. As a result, none of the three adverbs is likely to be proscribed.

• Moer. ε 73Moer. ε 73: εἴσκυψον, οὐκ ἔγκυψον· Ἀριστοφάνης (‘εἴσκυψον [‘stoop down and peep in’? ‘Pop in’?], not ἔγκυψον: Aristophanes [?]’; Hansen classifies this entry as a new fragment of Aristophanes). ἐγκύπτω is used by both Aristophanes (Nu. 191, Ra. 425, 804) and Plato (see Austin, Olson 2004, 132, according to whom the entry is ‘somewhat puzzling’), while εἰσκύπτω only occurs in LXX 1Re. 13.18.3, Teucer Cyz. FGrHist 274 F 3 (1st century BCE), Ios. AI 15.412.3, Eus. HE 10.4.22, and Thdt. MPG 81.100.10 (4th–5th century CE). I would not discount the possibility that here the structure x οὐκ y means that εἰσκύπτω is exceptionally used by Aristophanes in a (now lost) locus classicus instead of the otherwise usual ἐγκύπτω. This would be unparalleled in Moeris’ lexicon, but Valente (2015, 48–9) identifies and comments upon the same feature in the Antiatticist. Elsewhere too, Moeris discusses contextual uses in Aristophanes’ comedies: see Moer. φ 11Moer. φ 11, where Moeris understands φαῦλος and φλαῦρος as occasionally being used synonymously in Aristophanes.

• Moer. μ 31Moer. μ 31: μεριμνῶν ἀντὶ τοῦ φροντίζων Ξενοφῶν (‘Xenophon [Mem. 1.1.14] [said] μεριμνῶν [‘meditating upon’, pres. ptp.] instead of φροντίζων’). μεριμνῶν is not likely to be an approved form from the point of view of a strict Atticist. The comparative rarity of μεριμνάω, together with its ‘philosophical’ specialisation and occurrences in tragedy and lyric poetry (Willi 2003, 106; Beroutsos 2005, 68; Bagordo 2017, 45 on Ar. fr. 691; see also Finglass 2018, 507), may have led Moeris to reject it, with its use by Xenophon being insufficient to justify it. On the contrary, φροντίζω is widespread in the work of canonical Attic authors such as Aristophanes and Plato.

In conclusion, words and expressions found in canonical authors are not always recommended, especially when they are used only sparingly. Yet the opposite may also occur, whereby an unusual expression is recommended on the basis that it is found in a major authority. Then again, it seems that not all of the entries citing authors are in fact prescriptive in nature. This may at least partly depend on the different sources consulted by Moeris, but later additions and abridgments must have significantly contributed to the shaping of the lexicon as we know it.

F. Evaluative terminology

As noted above, the contraposition between Ἀττικοί, Ἕλληνες, and κοινόν is typical of Moeris’ lexicon. Admittedly, it is very difficult to give a clear-cut definition of these labels and their mutual relations. Indeed, linguistic analysis, although necessary, may not always be sufficient for understanding Moeris’ prescriptions. In doing so, one should also take into account a number of variables, including the sources consulted, the extremely limited corpus of approved authors, the author’s supposed will to remark on an identity in opposition to the ‘multitude’ of speakers (see e.g. the remarks in Swain 1996, 27–42; Matthaios 2013, 105–14), possible accretions and/or epitomisations of the lexicon, accidents in the manuscript tradition, and more. As I shall demonstrate in the following, all of these variables may influence our understanding of the evaluative terminology used by Moeris.

Drawing contrasts between Attic-speakers and Greek-speakers is not unique to Moeris. With the exception of a fragment from the 3rd-century BCE comic poet Posidippus (fr. 30.2–3: σὺ μὲν ἀττικίζεις, […] οἱ δ’ Ἕλληνες ἑλληνίζομεν ‘you speak Attic, while we, the Greeks, speak Greek’), the closest parallel is perhaps P.Oxy. 7.1012v, frr. 16–7 (= TM 64229) [205–250 CE], which flags differences in semantics and accentuation between Attic-speakers and Greek-speakers (see Ucciardello 2012, 79 n. 5). Furthermore, the distinction between Ionian-speakers and Greek-speakers repeatedly found in Aelius Dionysius may have been a model for the bipartite structure of most entries in Moeris’ lexicon, as Erbse (1950, 57) argues.

Recommended forms are usually referred to with the label Ἀττικοίοἱ Ἀττικοί (the adverb Ἀττικῶς is found five times, only in manuscripts belonging to the Vatican redaction). The label παλαιοίοἱ παλαιοί, found only in Moer. α 154Moer. α 154, ε 6Moer. ε 6, ε 7Moer. ε 7, ε 22Moer. ε 22, τ 21Moer. τ 21, and φ 16Moer. φ 16 has the same meaning of Ἀττικοί. Moeris’ classification, however, might not always have been limited to contrasting ancient and contemporary speakers, and there may be more nuances in this terminology. Moer. δ 29Moer. δ 29, ζ 8Moer. ζ 8 (see entry ἀπολλύασιν, ἀπολλύουσιν, and other 3rd person plurals of -νύμι verbs), and π 79Moer. π 79 draw a distinction between ‘old Attic’Attic, old and ‘second Attic’Attic, later. A two-way distinction between an older and a later stage of the Attic dialect may be traced back to Hellenistic scholarship: AristarchusAristarchus probably considered Homer as a representative of old Attic: see Galligani (2001, 44–6); Schironi (2018, 167–85); Montana (2020, 214); the entry ἀπολλύασιν, ἀπολλύουσιν, and other 3rd person plurals of -νύμι verbs; cf. also Probert (2004, 288–90). This distinction is comparable with the one attested, among Atticist lexica, in Phryn. Ecl. 390Phryn. Ecl. 390: πορνοκόπος· οὕτω Μένανδρος, οἱ δ’ ἀρχαῖοι Ἀθηναῖοι πορνότριψ λέγουσιν (‘πορνοκόπος [‘one who has commerce with prostitutes’]: So Menander [fr. 585]. Ancient Athenians say πορνότριψ’); see entry πορνοκόπος, πορνοκοπέω. Phryn. Ecl. 391Phryn. Ecl. 391: λήθαργος· οὕτω Μένανδρος, οἱ δ’ ἀρχαῖοι Ἀθηναῖοι ἐπιλήσμονα καλοῦσιν, οἷς πειστέον (‘λήθαργος [‘forgetful’]: So Menander [fr. 586]. Ancient Athenians say ἐπιλήσμων: one must obey them’); see entry λήθαργος, ἐπιλήσμων. Cf. perhaps also Phryn. Ecl. 177Phryn. Ecl. 177.

Moer. χ 12Moer. χ 12, however, distinguishes between ‘old Attic’Attic, old and ‘middle Attic’Attic, middle (see entry χολάς, χόλιξ), thus revealing a tripartite distinction that can be traced back at least to Aelius Dionysius (pace Wendel 1932, 2504, according to whom it may be due to Moeris himself; in Ar. fr. 706 the tripartite distinction of the Attic dialect has a socio-linguistic connotation which is also found in some Byzantine treatises on dialects; see Schoubben et al. 2023, 936–7): see Eust. in Od. 2.74.26–8: ἔα […] διαλελύσθαι δοκεῖ ἀπὸ τοῦ ἦ, οὗ μνήμη καὶ παρὰ Αἰλίῳ Διονυσίῳ, εἰπόντι ὡς Ἴωνες μὲν ἔα φασίν, Ἀττικοὶ δὲ ἦ μονοσυλλάβως, οἱ δὲ μέσοι σὺν τῷ νυ, οἷον, ‘ἐπειδὴ μεστὸς ἦν, ἀνεπαυόμην’ (‘ἔα […] is apparently a resolution of ἦ [‘I was’, impf. ind.], which is also mentioned by Aelius Dionysius [η 1]Ael.Dion. η 1: he states that Ionians say ἔα, while users of Attic [say] ἦ, with a monosyllable – anyway, users of middle [Attic say it] with ν, like ‘I didn’t stop till I was [ἦν] full’ [Ar. Pl. 695; transl. Henderson 2002, 527]’), where Aristophanes is considered a middle-Attic author, just as in Moer. χ 12. We cannot, therefore, exclude the possibility that the lexicon presents two different theories, depending on the sources consulted. The picture is further complicated by the possible role played by epitomisation: Probert (2004, 287), for example, notes that identical labels sometimes appear to be used in reference to both old Attic and later Attic.

The label ἝλληνεςἝλληνες mostly refers to the written koine of Moeris’ own times, as Maidhof (1912, 319–38) states (see also Swain 1996, 52; Monaco 2024, 41–3). Ἑλληνικόν Ἑλληνικός occurs four times with the same meaning. The forms marked as ‘Greek’ in Moeris’ lexicon very often ‘represent the kinds of morphological, phonetic and syntactic phenomena that characterize early and middle Post-classical Greek (3rd century BCE–3rd century CE)’ (Monaco 2024, 41). In this regard, note that ἡμεῖςἡμεῖς (Moer. δ 25Moer. δ 25, σ 20Moer. σ 20, ψ 12Moer. ψ 12), ἡμεῖς λέγομεν (Moer. π 30Moer. π 30), and καλοῦμεν (Moer. π 47Moer. π 47) may all refer to the same category of speakers that tend to be called Ἕλληνες (Swain 1996, 52; Monaco 2024, 42–3).

Nevertheless, here and there Ἕλληνες is also applied to forms that are perceived to be the standard, ‘unmarked’ forms, unlike those attributed to Attic-speakers. We see this clearly in Moer. λ 1Moer. λ 1, for example (see entry αὐτοσχεδιάζειν, ἥκειν, λαμβάνειν). Moreover, Moer. ι 9Moer. ι 9 and λ 2Moer. λ 2 seem to use Ἕλληνες to mean ‘correct use’ (in Moer. ε 3Moer. ε 3 Ἕλληνες is not actually found in the MSS, despite Hansen’s text: see Sandri 2023, 138; in Moer. σ 18Moer. σ 18, Hansen corrects what is given as Ἕλληνες in the text of cod. C to read Ἀττικοί, based on a note in cod. F and Phot. ε 2264; cf. also Thom.Mag. 146.2). This may depend on the sources consulted, which must have referred to the criteria of Ἑλληνισμός, i.e. ‘language correctness’ (see e.g. Pagani 2015, with further bibliography). The use of Ἕλληνες to mean ‘correct use’ is the exception in Moeris’ lexicon, and it is conceivable that the aforementioned entries have escaped a homogenisation process. Indeed, in other cases, forms marked as ‘Greek’ (= ‘correct’) in one of Moeris’ sources, may have been newly labelled as ‘Attic’: see e.g. Moer. τ 7Moer. τ 7: τάχιον οὐ λέγεται παρ’ Ἀττικοῖς ἀλλὰ θᾶττον (‘Users of Attic do not say τάχιον [‘sooner’], but θᾶττον’), probably depending on Phryn. Ecl. 52Phryn. Ecl. 52: τάχιον Ἕλληνες οὐ λέγουσιν, θᾶττον δέ (‘Users of Greek do not say τάχιον, but θᾶττον’; see entry θάττων, ταχίων. For such a use of Ἕλληνες in Phrynichus, see also Phryn. Ecl. 73Phryn. Ecl. 73).

The term κοινόνκοινός has long been the subject of debate. Jannaris (1903, 95 n. 1), for instance, interpreted it as a marker indicating ‘the stock common to all dialects including Attic, a sort of panhellenic Greek’ (for a similar view in previous scholarship, see Maidhof 1912, 292). According to Maidhof’s still influential reading (Maidhof 1912, 291–319; cf. e.g. Swain 1996, 52–3), however, κοινόν mostly refers to everyday language, the lower spoken koine of Moeris’ own times. Monaco (2024)    offers a new, detailed analysis of the entries in which κοινόν appears and argues that ‘common’ often indicates an unmarked form pertaining to ‘an ordinary type of Greek that could be used in any register’ (Monaco 2024, 69). These definitions are hardly reconcilable, and it is tempting to conclude that Moeris’ appraisals are somewhat inconsistent (on this point, see the bibliography in Maidhof 1912, 319–20). The following presents a number of case studies that illustrate the variety of meanings κοινόν could assume in Moeris’ lexicon. Before analysing the single entries, however, it is worth commenting on methodology. The Vatican redaction tends to substitute Ἕλληνες or Ἑλληνικὸν καὶ κοινόν with simple κοινόν. See the following list (here, I tacitly correct some mistakes found in Hansen’s apparatus):

Moer. α 18 καινότερον Ἀττικοί καὶ Ἕλληνες C : κοινότερον V : κοινόν F || α 58 ἀνοησίαν Ἕλληνες C : ἀνοησία κοινόν F : ἀνοητία· ἀνοησία D¹ || α 61 Ἕλληνες CFD¹ : κοινόν V || α 96 ἑλληνικὸν καὶ κοινόν C : κοινόν VD¹D² : om. F || β 16 Ἕλληνες C : κοινόν F || β 40 Ἕλληνες C : κοινόν F || γ 15 ἑλληνικὸν καὶ κοινόν C : κοινόν VD² : om. F || γ 17 Ἕλληνες C : κοινόν V || δ 12 (CVFD²) κοινόν … Πλάτωνι om. F || ε 58 Ἕλληνες καὶ κοινόν C : κοινόν VFD²E || η 12 Ἕλληνες CVE : κοινῶς F || η 20 Ἕλληνες C : κοινῶς F || θ 11 Ἀττικοί … Ἕλληνες V : Ἕλληνες om. C : ἀττικῶς … κοινῶς F^pc (κοινῶς … ἀττικῶς F^ac) || θ 12 Ἕλληνες CV : κοινῶς F || ο 28 (CVF) κοινόν supra Ἕλληνες F || π 6 Ἕλληνες CV : κοινῶς F || σ 31 Ἕλληνες C : κοινόν F : om. V || σ 33 Ἕλληνες CV : κοινόν F || τ 13 Ἕλληνες C : κοινόν V : om. F.

This tendency is particularly pronounced in cod. F, which repeatedly uses κοινόν even where codd. C and V agree in their readings of Ἕλληνες. It is also detectable in some entries of Thomas MagisterThomas Magister’s lexicon, which takes Moeris for its source, and where certain entries are derived from a copy of Moeris’ lexicon that belongs to the Vatican redaction (see Hansen 1998, 30–1; Monaco 2024, 48 n. 73, who compares e.g. Thom.Mag. 332.2 and Moer. σ 33Moer. σ 33, Thom.Mag. 197.4 and Moer. κ 46Moer. κ 46, Thom.Mag. 199.8 and Moer. κ 64Moer. κ 64, Thom.Mag. 278.10 and Moer. π 44Moer. π 44, and Thom.Mag. 146.3 and Moer. η 13Moer. η 13). Such considerations may also be useful for the constitutio textus. In Moer. ε 58Moer. ε 58, for instance, Hansen prints κοινόν, as is found in the Vatican redaction (codd. VFD²E). Nonetheless, in light of what we have just seen, Ἕλληνες καὶ κοινόν – as found in cod. C – is probably the best choice, since it seemingly preserves the unabridged text (cf. Moer. α 96Moer. α 96 and Moer. γ 15Moer. γ 15). In conclusion, it is better to leave aside those cases in which κοινόν is found in entries that only feature in the Vatican redaction.

Even when κοινόν is found in both redactions, however, it does not have the same meaning across these iterations. See the following examples:

• Moer. σ 38Moer. σ 38: σμικρόν Ἀττικοί· μικρόν κοινόν (‘Users of Attic [say] σμικρόν. μικρόν [is] common’). Here, κοινόν does not appear to have any particular connotations, but seems to mean ‘unmarked’ – a meaning elsewhere expressed by Ἕλληνες, as we have seen above. Monaco (2024, 53–4) explains this and other similar cases, arguing that κοινόν refers to a non-connoted form that fits every register of Greek. This may well be true. Nevertheless, we cannot exclude the possibility that a longer evaluation may have been lost during transmission. In particular, κοινόν may be an abbreviated form of Ἕλληνες καὶ κοινόν: if so, we should also conclude that the process of shortening that is evident in the Vatican redaction had already begun before the archetype of both the Parisian and Vatican redactions had been copied. Another possible explanation is that here κοινόν has the contextual meaning of ‘common to Attic and other dialects’ or ‘common Greek (from which individual dialects are derived through linguistic changes)’, i.e. it corresponds to one of the meanings assumed by κοινός and cognates in the coeval grammatical works (a survey in Maidhof 1912, 281–90; see also Consani 1991, 15–53; similar considerations may be made for Moer. β 13Moer. β 13: βοῦς ἑνικῶς καὶ πληθυντικῶς <Ἀττικοί>· βόες κοινόν, ‘<Users of Attic> [say] βοῦς [‘bull’, ‘cow’, ‘cattle’] both in the singular and in the plural. βόες [nom. pl.] is common’; yet the entry is only in codd. FE and should be taken with a pinch of salt).

• Moer. ε 60Moer. ε 60: εἰ γάρ Ἀττικοί· εἴθε γάρ κοινόν (‘Users of Attic [say] εἰ γάρ [‘if only’]. εἴθε γάρ [is] common’). This entry seems to contradict Maidhof’s conclusions. Indeed, according to La Roi (2022, 218), εἴθε γάρ ‘is only found in high register contexts in early Post-Classical Greek […]. At the same time, it is found in middle Post-Classical Greek in contexts marked by a register upgrade, such as to represent older language in histories […] or once in a high register official papyrus’. Here, however, cod. V reads Ἕλληνες (κοινόν C : om. F), which may be the correct reading in light of La Roi’s aforementioned considerations.

• Moer. π 49Moer. π 49: ποῖ Ἀττικοί· ποῦ κοινόν (‘Users of Attic [say] ποῖ [‘to what place?’]. ποῦ [lit. ‘in what place?’] [is] common’). On this entry, see Maidhof (1912, 298–300). Here I only wish to remark that the label κοινόν probably qualifies a use felt to be tout court incorrect; see Phryn. Ecl. 28Phryn. Ecl. 28: ποῖ ἄπει· οὕτω συντάσσεται διὰ τοῦ ι· ποῦ δὲ ἄπει, διὰ τοῦ υ, ἁμάρτημα. εἰ δὲ ἐν τῷ υ, ποῦ διατρίβεις (‘ποῖ ἄπει [‘to what place will you go?’]: It is constructed in this way, with ι: ποῦ δὲ ἄπει, with υ, is a mistake. If [the adverb ends] in υ, [the correct construction is] ποῦ διατρίβεις [‘where do you spend your time?’]’).

• Moer. κ 49Moer. κ 49: κάθησο Ἀττικοί· κάθου κοινόν (‘Users of Attic [say] κάθησο [‘sit down’, imper.]. κάθου [is] common’). κάθου is well attested in the koine, having many occurrences, for example, in the Septuagint (see Maidhof 1912, 300–2). Various erudite sources, however, present both κάθησο and κάθου as Attic: in particular, milder Atticist sources defend the form κάθου: see the apparatus in Valente (2015, 187); Arnott (1996, 647); Bagordo (2016, 158). This may depend on the fact that κάθησο and κάθου were respectively regarded as ‘old’ and ‘new’ Attic forms (even if we know that contracted forms occur in Attic tragedy and in Old Comedy, as well: see the bibliography in Monaco 2024, 58). Indeed, κάθησο is already found in Homer (who was an Athenian according to some ancient scholars, see E. Canon), while κάθου is found in Middle and New Comedy, e.g. Eust. in Od. 2.171.7–8 says: τοῦ δὲ ἧσο σύνθετον τὸ κάθησο, ὅθεν ἐνδείᾳ, φασί, καὶ συστολῇ κάθεο, καὶ συναλιφῇ κάθου παρὰ Μενάνδρῳ (‘κάθησο is a compound of ἧσο, from which, as someone says, [one has] κάθεο through defect [of σ] and change [of η into ε], and κάθου, through coalescing [of the two syllables into one], in Menander [fr. 475]’). Moeris’ proscription of κάθου goes in the opposite direction compared to other Atticist sources, which attempt to reconstruct an authoritative basis for κάθου. This could also be the case with those entries in Moeris’ lexicon in which a form is simultaneously ascribed to an Attic author (e.g. Demosthenes) and rejected as κοινόν. Moeris is, however, not always consistent in this regard and his position probably depends on the different sources consulted: unlike in the present case, Moer. α 32Moer. α 32, ε 65Moer. ε 65, and η 22Moer. η 22 labels non-contracted forms as Greek, while the contracted ones are said to be Attic.

• Moer. γ 10Moer. γ 10: γλαμῶσα Ἀττικοί· λημῶσα κοινὸν †ἀμφότερον† (‘Users of Attic [say] γλαμῶσα [‘being blear-eyed’, pres. pt. fem.]; λημῶσα [is] common †both [forms]/in both ways†’; the reading of cod. C – which according to Sandri 2023, 137 should not be maintained in the text – is actually ἀμφότερον, and not ἀμφότερα, printed by Hansen, see already Maidhof 1912, 313 n. 1; ἀμφότερον is missing in V). At a first glance, Moeris’ proscription of λημάω as ‘common’ seems odd: the verb, already attested in Aristophanes (see below), is apparently prescribed in Pollux (2.65Poll. 2.65; 4.185Poll. 4.185) and is used by Lucian (6x). Moreover, Phot. γ 121 says that λημᾶν is ‘Greek’, i.e. ‘correct’ (cf. also Phot. λ 262). We cannot, however, exclude the possibility that Moeris’ entry argues against the admissibility of the participial forms of λημάω (despite Ar. Pl. 581 λημῶντες) and not the verb itself: indeed, the participle γλαμῶν is apparently preferred to λημῶν by Philemo (Vindob.) 395.10–1Philemo (Vindob.) 395.10–1: λημᾷς σύ γ’, οὐ γλημᾷς <Ἀριστοφάνης λέγει>, λημῶντα δ’ οὐκέτ’, ἀλλὰ <σύ γε> γλαμῶντ’ <ἐρεῖς> (‘<Aristophanes says> λημᾷς σύ [‘you are blear-eyed’, cf. Ar. Nu. 327], not γλημᾷς. Nevertheless, <you shall> not <say> λημῶντα, but γλαμῶντα’; Philemo [Laur.] 357Philemo (Laur.) 357 γλαμμᾶν, ἀλλ’ οὐ λημμᾶν, ‘[Say] γλαμμᾶν [‘to be blear-eyed’], but not λημμᾶν’, seems to be heavily epitomised and does not help to shed light on the problem). Admittedly, this is a tentative explanation. Yet a participle γλαμῶν was perhaps found as a varia lectio in Ar. Ra. 588 (codd. read γλάμων): cf. Hsch. γ 593: *γλαμῶν· λημῶν (pΣ), which may depend on Atticist sources; and see schol. Ar. Ra. 588a, according to which the grammarian Callistratus (2nd century BCE) compared γλάμων with χάρων, thus probably indicating the inflectional model to which γλάμων was to be referred. The very need to establish a comparison between γλάμων and χάρων suggests that the reading γλάμων found in Ar. Ra. 588 was not assured and an alternative (γλαμῶν) was also possible.

• Moer. ι 15Moer. ι 15: ἴτρια· πλάσματα λεπτὰ σησαμῇ πεπλασμένα· λάγγανα κοινόν (‘ἴτρια: Small cakes made with roasted sesame-seeds and pounded with honey. λάγγανα is common’; on this entry, see Maidhof 1912, 310–1). One wonders whether Moeris knew of the reading λάγανα found in Ar. Ec. 843 (codd. unanimously read πόπανα), testified by Ath. 3.110a in a section dealing with λάγανα. Be that as it may, λάγανα is probably rejected as κοινόν because it was in current use among Moeris’ contemporaries; see Gal. 6.768.13–5: ὀνομάζειν δέ μοι δοκοῦσιν ταῦτα τὰ νῦν ὑφ’ ἡμῶν καλούμενα λάγανά τε καὶ ῥυήματα κοινῇ προσηγορίᾳ τῇ τῶν ἰτρίων οἱ παλαιοί (‘It seems to me that ancient [authors] refer to those which we currently call λάγανα and ῥυήματα by using the general expression ἴτρια’; cf. also Gal. 6.492.4–5).

• Moer. φ 14Moer. φ 14: φειδωλοί Ἀττικοί· σκνιφοί κοινόν (‘Users of Attic [say] φειδωλοί [‘sparing’, ‘niggard’, nom. pl.]. σκνιφοί [is] common’; on this entry, see Maidhof 1912, 296–8). Phryn. Ecl. 377Phryn. Ecl. 377 ascribes the form σκνιφός to the ‘multitude’ (οἱ πολλοίοἱ πολλοί), a label that is a derogatory way of referring to contemporary speakers (unlike in Pollux: see Strobel 2011, 137; Matthaios 2013, 95–105; Monaco 2024, 62–3).

• Moer. α 151Moer. α 151: ἄμυναν ἡ κοινὴ συνήθεια· λέγει δὲ τῶν Ἀττικῶν οὐδείς (‘The contemporary use [says] ἄμυνα [‘self-defence’, ‘vengeance’]: no Attic-speaker says so’; on this entry, see also C. Content and structure). Here, κοινὴ συνήθειασυνήθεια probably refers to ‘contemporary use’; see e.g. Valente (2013, 150), with further bibliography. It may well be that, in other cases, κοινόν was simply a substitute for the expression κοινὴ συνήθεια, which Moeris found in his sources. This would also partly explain why some uses labelled as κοινόν, although rejected, are associated with Attic authors: indeed, in Herodian’s words (Περὶ μονήρους λέξεως GG 3,2.910.10–2), ‘everyday usage […] sometimes knows a use of words just like the ancient Greeks’ (transl. Sluiter 2011, 304; cf. Valente 2013, 152–3). Unlike both the Antiatticist and Pollux, however, Moeris usually condemns contemporary forms even when they are credited with an authoritative pedigree.

The picture that emerges from the examples above is doubtless marked by variety. And as if that were not enough, we can detect further complications here and there. In Moer. φ 22Moer. φ 22, we find: φθεῖρες ἀρσενικῶς Ἀττικοί· θηλυκῶς Ἕλληνες (‘Users of Attic [employ] φθεῖρες [‘lauses’] in the masculine. Users of Greek [employ it] in the feminine’), while Thom.Mag. 377.2–5 gives: ὁ φθείρ Ἀττικόν, ἡ φθείρ δὲ κοινόν (‘ὁ φθείρ is Attic, while ἡ φθείρ is common’); in Choerob. in Theodos. GG 4,1.321.13–5 (= Hdn. Περὶ κλίσεως ὀνομάτων GG 3,2.748.13–4), we find: φθείρ φθειρός· τοῦτο δὲ ἀρσενικῶς λέγεται, οὐδὲ γάρ, ὡς λέγουσί τινες, τὰς φθεῖρας θηλυκῶς ἐν τῇ συνηθείᾳ (‘φθείρ, φθειρός: It is masculine, and not feminine – τὰς φθεῖρας –, as someone says in everyday language’); and schol. rec. Ar. Pl. 537a reads: φθειρῶν· αἱ κοινῶς λεγόμεναι φθεῖραι (Cang) (‘φθειρῶν: Those commonly called φθεῖραι’). As we have seen above, forms ascribed to everyday language in other sources are sometimes classified as ‘common’ in Moeris’ lexicon. Indeed, it is possible that Moeris originally classified the feminine form as ‘common’. In this context, it is worth pointing to Moer. μ 14Moer. μ 14: μασχάλη Ἀττικοί· μάλη Ἕλληνες (‘Users of Attic [say] μασχάλη [‘arm-pit’]. Users of Greek [say] μάλη’). This can be compared with Poll. 2.139Poll. 2.139: ἡ δὲ μασχάλη ὑπὸ μὲν τῶν ἰδιωτῶν καλεῖται μάλη, ὑπὸ δὲ τῶν Ἀττικῶν οὐχί (‘The μασχάλη is called μάλη by laymen, but not by Attic-speakers’); cf. also Phryn. Ecl. 169Phryn. Ecl. 169, Phot. υ 223, Phot. υ 224, and see Matthaios (2013, 109–10). The label ἰδιῶταιοἱ ἰδιῶται in Pollux refers to the everyday language of uncultivated speakers (see Matthaios 2013, 105–14): surprisingly, Pollux’ evaluation seems even stricter than that of Moeris, unless we think that Moeris originally classified μάλη as ‘common’.

As we have seen above, Moeris’ lexicon provides no single definition of κοινόν. It can indicate a form found in the everyday language of contemporary speakers (similarly to what other sources call κοινὴ συνήθεια, οἱ πολλοί, etc.). It can also refer to a form common to both cultivated koine and contemporary everyday language (i.e. probably the same as those forms labelled with Ἕλληνες καὶ κοινόν, on which see entry ἄθλιος, entry ἀμείνω, ἥττω, entry γόης: the confusion may be partially due to processes of epitomisation). Furthermore, sometimes κοινόν is apparently referred to incorrect uses. Finally, it may also simply mean ‘unmarked’ (even if it is not clear in what sense a ‘common’ form should be unmarked in Moeris’ perspective: see what has been noted above about κοινόν in other coeval erudite works).

In Moer. δ 48Moer. δ 48, μ 11Moer. μ 11, and μ 17Moer. μ 17 the forms ascribed to Greek-speakers are said to be ‘analogical’, i.e. regular according to the inflectional rules devised by ancient grammarians on the basis of the criterion of analogyAnalogy (see entry Δημοσθένας, Δημοσθένεις, entry μονομάχης, μονομάχος, entry Σωκράτη, Σώκρατες). This seemingly implies a conception of dialectal forms as being derived from standard ones through linguistic changes (that is, what ancient grammarians called πάθηπάθη). Undoubtedly, and as Wendel (1932, 2504) argues, the sporadic description of a rejected form as ‘regular’ only reinforces the impression that, for the large part, Moeris’ view is that of a strict analogist; indeed, the same has often been said of one of his sources, Phrynichus (see Alpers 1981, 100; Tosi 1988, 176–95; Sonnino 2014, 169 n. 21). Then again, Moer. χ 6Moer. χ 6, for example, recommends χθές and χθιζόν, while the ‘analogical’ ἐχθές and ἐχθεσινόν are rejected as ‘Greek’ (see the erudite sources cited in the apparatus of Orus fr. B 73 in Alpers 1981, 223). And there are further cases similar to this one: Moer. α 77Moer. α 77, for instance, prescribes ἀσκαρίζειν (‘to jump, throb, palpitate’) instead of σκαρίζειν, possibly regarding the latter as the regular form from which ἀσκαρίζειν was created ‘by addition of a prothetic vowel’, as Phryn. PS 42.7Phryn. PS 42.7 says; and Moer. θ 14Moer. θ 14 prescribes θράττει (‘he/she/it troubles, disturbs’) while rejecting ταράσσει, seemingly following Phryn. PS 75.6Phryn. PS 75.6, according to whom θράττω derives from ταράσσω through syncope and change of τ into θ.

On the isolated καινότερον (C : κοινότερον V : κοινόν F) in Moer. α 18Moer. α 18, see Gerbi’s (forthcoming) cautious considerations on the topic.